Add support for optical marker-based motion capture data

[sfmig]

Is your feature request related to a problem? Please describe.
movement was initially conceived as a 2D and 3D data analysis tool. The gold-standard for 3D pose data is currently marker-based. It would be nice to support common file formats in marker-based motion capture in movement. This would further expand our user base, and give us access to quite good quality datasets.

Describe the solution you'd like
I think the two following are the most popular formats (both provide an example file for reference):

• c3d
• bvh - this one actually describes the pose of a skeleton, usually computed from markers placed in specific points (using marker-suits for example). We could use this Python package bvhio to extract the world coordinates of the joints and treat them as movement keypoints.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context

• There are very good and clean datasets collected with marker-based motion capture (e.g., see dog dataset here), so it could be useful to do sanity checks on real data when implementing 3D functionality.
• It might be useful to check the kinetics toolkit as reference

[niksirbi]

Also related to #124, which is about loading 3D data from triangulated marker-less pose tracking.